Tech

The next AI frontier – the Deepfakes are calling

Published

3 years ago

March 18, 2023

Voice deepfakes are calling – here’s what they are and how to avoid getting scammed

You have just returned home after a long day at work and are about to sit down for dinner when suddenly your phone starts buzzing. On the other end is a loved one, perhaps a parent, a child or a childhood friend, begging you to send them money immediately.

You ask them questions, attempting to understand. There is something off about their answers, which are either vague or out of character, and sometimes there is a peculiar delay, almost as though they were thinking a little too slowly. Yet, you are certain that it is definitely your loved one speaking: That is their voice you hear, and the caller ID is showing their number. Chalking up the strangeness to their panic, you dutifully send the money to the bank account they provide you.

The next day, you call them back to make sure everything is all right. Your loved one has no idea what you are talking about. That is because they never called you – you have been tricked by technology: a voice deepfake. Thousands of people were scammed this way in 2022.

As computer security researchers, we see that ongoing advancements in deep-learning algorithms, audio editing and engineering, and synthetic voice generation have meant that it is increasingly possible to convincingly simulate a person’s voice.

Even worse, chatbots like ChatGPT are starting to generate realistic scripts with adaptive real-time responses. By combining these technologies with voice generation, a deepfake goes from being a static recording to a live, lifelike avatar that can convincingly have a phone conversation.

Crafting a compelling high-quality deepfake, whether video or audio, is not the easiest thing to do. It requires a wealth of artistic and technical skills, powerful hardware and a fairly hefty sample of the target voice.

There are a growing number of services offering to produce moderate- to high-quality voice clones for a fee, and some voice deepfake tools need a sample of only a minute long, or even just a few seconds, to produce a voice clone that could be convincing enough to fool someone. However, to convince a loved one – for example, to use in an impersonation scam – it would likely take a significantly larger sample.

With all that said, we at the DeFake Project of the Rochester Institute of Technology, the University of Mississippi and Michigan State University, and other researchers are working hard to be able to detect video and audio deepfakes and limit the harm they cause. There are also straightforward and everyday actions that you can take to protect yourself.

For starters, voice phishing, or “vishing,” scams like the one described above are the most likely voice deepfakes you might encounter in everyday life, both at work and at home. In 2019, an energy firm was scammed out of US$243,000 when criminals simulated the voice of its parent company’s boss to order an employee to transfer funds to a supplier. In 2022, people were swindled out of an estimated $11 million by simulated voices, including of close, personal connections.

What can you do?

Be mindful of unexpected calls, even from people you know well. This is not to say you need to schedule every call, but it helps to at least email or text message ahead. Also, do not rely on caller ID, since that can be faked, too. For example, if you receive a call from someone claiming to represent your bank, hang up and call the bank directly to confirm the call’s legitimacy. Be sure to use the number you have written down, saved in your contacts list or that you can find on Google.

Additionally, be careful with your personal identifying information, like your Social Security number, home address, birth date, phone number, middle name and even the names of your children and pets. Scammers can use this information to impersonate you to banks, realtors and others, enriching themselves while bankrupting you or destroying your credit.

Here is another piece of advice: know yourself. Specifically, know your intellectual and emotional biases and vulnerabilities. This is good life advice in general, but it is key to protect yourself from being manipulated. Scammers typically seek to suss out and then prey on your financial anxieties, your political attachments or other inclinations, whatever those may be.

This alertness is also a decent defense against disinformation using voice deepfakes. Deepfakes can be used to take advantage of your confirmation bias, or what you are inclined to believe about someone.

If you hear an important person, whether from your community or the government, saying something that either seems very uncharacteristic for them or confirms your worst suspicions of them, you would be wise to be wary.

Tech

SoftBank plans acquisition of DigitalBridge for AI expansion

SoftBank advances towards acquiring DigitalBridge to boost AI infrastructure amid soaring global data center demand

Published

3 days ago

December 30, 2025

Harvey Vargas

SoftBank advances towards acquiring DigitalBridge to boost AI infrastructure amid soaring global data center demand

In Short:
– SoftBank may acquire DigitalBridge to enhance its AI infrastructure amid rising global data centre demand.
– The deal could control $108 billion in digital assets, with financial details yet to be disclosed.

SoftBank Group is reportedly in advanced talks to acquire DigitalBridge Group, a move that would dramatically expand the Japanese conglomerate’s control over critical AI infrastructure as global demand for data centres accelerates. The potential deal, which could be announced within days, would give SoftBank exposure to roughly $108 billion in digital infrastructure assets, including data centres, cell towers and fibre networks. While financial terms remain undisclosed, the talks are said to be at an advanced stage.

The acquisition fits squarely into founder Masayoshi Son’s renewed bet on artificial intelligence and computing capacity. DigitalBridge manages investments in major data centre operators such as Vantage Data Centers, Switch, DataBank and AtlasEdge, placing SoftBank at the centre of the infrastructure powering next-generation AI. The company is also a key participant in Stargate, a $500 billion private-sector AI initiative announced earlier this year, and recently agreed to buy ABB’s robotics division as part of its broader push into physical AI.

Intensifying competition

Markets have reacted strongly to the prospect of the deal, with DigitalBridge shares surging as much as 47% after the initial reports emerged. The rally highlights intensifying competition for data centre assets, as AI drives unprecedented demand for computing power. McKinsey estimates AI-related infrastructure spending could reach $6.7 trillion by 2030, while Goldman Sachs forecasts global data centre power consumption will rise 175% from 2023 levels by the end of the decade. If completed, the acquisition would mark SoftBank’s return to direct ownership of a major digital infrastructure platform at a pivotal moment in the AI race.

Tech

Italy orders Meta to open WhatsApp to AI competitors

Italy orders Meta to allow rival AI chatbots on WhatsApp amid regulatory battle over market dominance

Published

4 days ago

December 29, 2025

Harvey Vargas

Italy orders Meta to allow rival AI chatbots on WhatsApp amid regulatory battle over market dominance

In Short:
– Italy’s antitrust authority requires Meta to allow access to rival AI chatbots on WhatsApp during an investigation.
– Meta plans to appeal the ruling, claiming it disrupts their system and questioning WhatsApp’s role as an AI service platform.

Italy’s antitrust authority has ordered Meta to allow competing AI chatbots access to WhatsApp, suspending rules that blocked rivals. The decision comes amid concerns that Meta’s policies could limit competition and harm consumers in the rapidly growing AI services market. Meta plans to appeal, calling the ruling “fundamentally flawed” and arguing that WhatsApp wasn’t designed to support third-party AI chatbots.

The Italian Competition Authority began investigating Meta after its March 2025 launch of Meta AI on WhatsApp, later expanding the probe to cover updated business terms that excluded rival AI providers, such as ChatGPT, Microsoft Copilot, and Perplexity. The European Commission has launched a parallel investigation, highlighting growing regulatory scrutiny on tech giants in Europe.

Europe’s stricter stance on Big Tech has sparked pushback from the industry and political figures in the U.S., including former President Donald Trump. Meta maintains that its Business API restrictions still allow AI for customer support and order tracking, but says general-purpose chatbot distribution falls outside its intended use.

Tech

China’s maglev breakthrough hits 700 km/h in seconds, reshaping the future of transport

China sets world record with maglev train hitting 700 km/h in just two seconds, revolutionising ultra-high-speed transport

Published

6 days ago

December 27, 2025

Harvey Vargas

China sets world record with maglev train hitting 700 km/h in just two seconds, revolutionising ultra-high-speed transport

In Short:
– Chinese researchers set a world record, accelerating a test vehicle to 700 km/h in two seconds.
– This milestone positions China as a leader in ultra-high-speed maglev technology and future transport developments.

China has set a new world record in magnetic levitation technology after accelerating a ton-class superconducting maglev test vehicle to 700 kilometres per hour in just two seconds. The achievement, reported by state broadcaster CCTV, marks the fastest acceleration ever recorded for an electric maglev system and cements China’s position at the forefront of ultra-high-speed transport innovation.

The test was conducted by researchers at the National University of Defense Technology on a 400-metre track, where footage showed the vehicle flashing across the rail-like structure in a blur, leaving a misty trail behind it. The breakthrough follows more than a decade of research tackling complex challenges such as ultra-high-speed electromagnetic propulsion, electric suspension guidance systems, and high-field superconducting magnets, all of which are critical to stable travel at extreme speeds.

Hyperloop technology

Beyond headline-grabbing velocity, the milestone opens the door to future transport systems, including vacuum-tube maglev networks, commonly referred to as hyperloop technology. Scientists say the same advancements could also be applied to aerospace launch assistance, electromagnetic launch systems, and advanced experimental testing. According to Professor Li Jie from the National University of Defense Technology, the successful trial will significantly accelerate China’s research into frontier technologies, with future work focusing on pipeline-based high-speed transport and aerospace equipment testing.

While China now leads in superconducting maglev acceleration, global competition remains fierce. Japan still holds the record for the fastest manned train, with its L0 Series maglev reaching 603 kilometres per hour during testing in 2015. China, however, operates the world’s only commercial maglev service — the Shanghai Maglev — which currently runs at 300 kilometres per hour after its top speed was reduced from 431 kilometres per hour in 2021.

The December test builds on earlier progress made this year, including a 1.1-ton test sled that reached 650 kilometres per hour in seven seconds over a 600-metre track in June 2025. Together, these developments signal rapid momentum in China’s push toward next-generation transport systems that could redefine how people and payloads move across the planet.